Web-plugin to Detect Clickbait in News Articles using RNN and LSTM

'Clickbaits' are either false or misleading headlines that aim to attract users' attention and entice them to click on a link or a news article that isn't really interesting as the title makes it up to be. A clickbait title can degrade the user experience and spread disinformation or misinformation on the Internet. In this study, a novel approach for Clickbait detection in news articles using machine learning technique of Recurrent Neural Network and Long-Short Term Memory is proposed. The clickbait detection process takes place in two distinct steps, first to identify the headlines in a website using a web plug-in, these headlines are then fed into a machine learning model which can predict if the headline is a clickbait or not. This a binary classification problem where the model is trying to predict which of the two categories the input belongs to. The algorithm is put into action using a web plugin which is triggered when popular news websites are visited and fetch the headlines using DOM method, the headlines are then sent to the server for prediction upon which the headlines get marked either red or green which indicates the headlines to be either clickbait or not respectively which, this makes the user aware if a headlineis clickbait or not with just a glance. The result obtained by the above method is an improvement over the existing methods and the user interface provided by the web plugin removes any manual labor from the user side and offers an engaging yetsimplistic user experience.